Met Office

Low‐level jets (LLJs) are sensitive to continental‐scale pressure gradients. Soil moisture influences these gradients by altering turbulent flux partitioning and near‐surface temperatures, thereby affecting LLJ characteristics. The Turkana jet, a strong southeasterly LLJ flowing through a channel between the Ethiopian and East African Highlands, is an important feature of the East African water cycle. Previous work has shown that the jet is sensitive to soil‐moisture‐induced pressure gradients driven by the Madden–Julian oscillation. Here, we build on this finding through using convection‐permitting UK Met Office Unified Model simulations to isolate the role of soil moisture in shaping jet characteristics. Modelling experiments reveal that the Turkana jet is highly sensitive to soil‐moisture‐induced temperature gradients across the channel's exit. Prescribing realistic dry soils intensifies the local surface‐induced thermal low and strengthens the jet. A maximum jet sensitivity of up to 8m·s−1$8\kern0.3em \mathrm{m}\cdotp {\mathrm{s}}^{-1}$ occurs when comparing dry and wet surface states within 750 km downstream of the exit, highlighting the significant influence of soil moisture on jet dynamics, given typical speeds of 8–14m·s−1$14\kern0.3em \mathrm{m}\cdotp {\mathrm{s}}^{-1}$. The impact of soil moisture on the jet is most pronounced when synoptic forcing is weak and skies are clear. Notably, despite a substantial impact on LLJ strength, we find a minor sensitivity of the vertically integrated moisture transport. We speculate that this minimal sensitivity is linked to model errors in the representation of boundary‐layer turbulence, which affects midtropospheric moisture and the strength of elevated nocturnal inversions. This study highlights that the Turkana channel is a hotspot for surface–jet interactions, due to the strong sensitivity of surface fluxes to soil moisture near a topographically constrained LLJ. Future research should continue examining surface‐driven predictability, particularly in regions where land–atmosphere interactions influence dynamical atmospheric conditions, and evaluate such processes in weather prediction models.

Current frameworks for evaluating biogeochemical climate change feedbacks in Earth System Models lack an explicit consideration of nitrogen cycling in the land and ocean spheres despite its vital role in limiting primary productivity. As coupled carbon-nitrogen cycling becomes the norm, a better understanding of the role of nitrogen cycling is needed. Here we develop a new framework for quantifying carbon-nitrogen feedbacks in Earth System Models and show that rising nitrogen deposition acts as a negative feedback over both land and ocean, enhancing carbon dioxide (CO2) fertilisation in a model ensemble. However, increased CO2 uptake due to rising nitrogen deposition is small relative to the large reduction in CO2 uptake when coupled carbon-nitrogen cycling is implemented in Earth System Models. Altogether, rising nitrogen deposition leads to only a minor increase in CO2 uptake but also enhances nitrous oxide (N2O) emissions over land and ocean, contributing only marginally to mitigating climate change.

Under the Paris Agreement, signatories aim to limit the global mean temperature increase to well below 2°C above pre‐industrial levels. To achieve this, many countries have made net zero greenhouse gas emissions targets, with the aim of halting global warming and stabilizing the climate. Here, we analyze the stability of global and local temperatures in an ensemble of simulations from the zero‐emissions commitment Model Intercomparison Project, where CO2 emissions are abruptly ceased. Our findings show that at both the global and local level stabilization does not occur immediately after net zero CO2 emissions. The multi‐model median (mean) global average temperature stabilizes after approximately 90 (124) years, with an inter‐model range of 64–330 years. However, for some models, this may underestimate the actual time to become stable, as this is the end of the simulation. Seven models exhibited cooling post‐emission cessation, with two of the models then warming after the initial cooling. One model gradually warmed through the entire simulation, while another had alternating cooling and warming. At the local level, responses varied significantly, with many models simulating the reversal of trends in some areas. Changes at the local level, at many locations, continue beyond the stabilization of global temperature and are not stable by the end of the simulations.

A quarter of the deforested Amazon has regrown as secondary tropical forest and yet the climatic importance of these complex regenerating landscapes is only beginning to be recognised. Advances in satellite remote-sensing have transformed our ability to detect and map changes in forest cover, while detailed ground-based measurements from permanent monitoring plots and eddy-covariance flux towers are providing new insights into the role of secondary forests in the climate system. This review summarises how progress in data availability on Amazonian secondary forests has led to better understanding of their influence on global, regional and local climate through carbon and non-carbon climate benefits. We discuss the climate implications of secondary forest disturbance and the progress in representing forest regrowth in climate models. Much remains to be learned about how secondary forests function and interact with climate, how these processes change with forest age, and the resilience of secondary forest ecosystems faced with increasing anthropogenic disturbance. Secondary forests face numerous threats: half of secondary forests in the Brazilian legal Amazon were 11 years old or younger in 2023. On average, 1%–2% of Amazon secondary forests burn each year, threatening the permanence of sequestered carbon. The forests that burn are predominantly young (in 2023, 55% of burned secondary forests were <6 years old, <4% were over 30 years old). In the context of legally binding international climate treaties and a rapidly changing political backdrop, we discuss the opportunities and challenges of encouraging tropical forest restoration to mitigate anthropogenic climate change. Amazon secondary forests could make a valuable contribution to Brazil’s Nationally Determined Contribution provided there are robust systems in place to ensure permanence. We consider how to improve communication between scientists and decision-makers and identify pressing areas of future research.

The North Atlantic Oscillation (NAO) dominates winters in Western Europe and eastern North America. Future climate model projections of the NAO are highly uncertain due to both modelled irreducible internal variability and different model responses. Here we show that some of the model spread in multi-decadal NAO simulations is caused by climatological water vapour errors, and develop an emergent constraint that reveals a substantial response of the NAO to volcanic eruptions and greenhouse gases (GHGs). Taking account of the signal-to-noise paradox apparent in these simulations suggests that under the high-emissions scenario the multi-decadal NAO will increase to unprecedented levels that will likely cause severe impacts, including increased flooding and storm damage. This can be avoided through mitigation to reduce GHG emissions. Our results suggest that taking model projections at face value and seeking consensus could leave society unprepared for impending extremes.

The UK experienced an unprecedented heatwave in 2022, with temperatures reaching 40 °C for the first time in recorded history. This extreme heat was accompanied by widespread fires across London and elsewhere in England, which destroyed houses and prompted evacuations. While attribution studies have identified a strong human fingerprint contributing to the heatwave, no studies have attributed the associated fires to anthropogenic influence. In this study, we assess the contribution of human-induced climate change to fire weather conditions over the summer of 2022 using simulations from the HadGEM3-A model with and without anthropogenic emissions and apply the Canadian Fire Weather Index. Our analysis reveals at least a 6-fold increase in the probability of very high fire weather in the UK due to human influence, most of which is driven by high fire conditions across England. These findings highlight the significant role of human-induced climate change in emerging UK wildfires. As we experience more hotter and drier summers as temperatures continue to rise the frequency and severity of fires are likely to increase, posing significant risks to both natural ecosystems and human populations. This study underscores the need for further research to quantify the changing fire risk due to our changing climate and the urgent requirement for mitigation and adaptation efforts to address the growing wildfire threat in the UK.

Despite the rapid development of global‐storm‐resolving models, computational expense hampers widespread deployment of these for operational forecasting. Thus, regional models at convection‐permitting resolutions still need to be deployed to take advantage of explicitly representing smaller‐scale processes that improve the forecast. Often, the choice of domain size is made subjectively, despite both domain size and location with respect to prevailing meteorology significantly impacting how constrained the regional model is to its driving model. This has implications on error growth, upscale impacts of mesoscale variability, along with ensemble spread. Here, we introduce a novel diagnostic designed to characterise lateral boundary spin‐up by quantifying the age of air, or the time since the air entered the model through the lateral boundaries. We apply this diagnostic to a variety of case studies over regional domains in Australia, and contrast this to a larger pan‐Australia domain, demonstrating that larger‐domain models exhibit more realistic atmospheric structures and reduced spin‐up at the boundaries, directly correlated with the age‐of‐air metric. Additionally, we show that large‐domain models demonstrate quicker spin‐up of mesoscale kinetic energy and subsequent upscale energy growth, evidenced by power spectral analysis. We further generalise the age‐of‐air diagnostic by applying it to reanalysis, providing climatological perspectives of the age of air which aid future applications such as characterising error growth in regional versus driving models in different weather regimes, and determining optimal time scales to blend regional and global model fields in data assimilation.

The inter-model difference in the tropical Pacific SST warming pattern is a big stumbling block for reliable projections of global climate change. Here by conducting an inter-model Empirical Orthogonal Function (EOF) analysis as well as an ocean mixed-layer heat budget, we find that the first two modes of inter-model difference in the SST warming pattern projected by 30 CMIP6 models, explaining more than three-quarters of the total inter-model variance, are both tied to different cloud–radiation feedbacks. The EOF1 mode that captures the different magnitudes of El Niño-like warming as well as the largest inter-model variance in the far eastern equatorial Pacific, is likely driven by highly diverse cloud–radiation feedbacks in the east and, to a lesser extent, by differing changes in the oceanic vertical temperature gradient. The EOF2 mode that mainly represents the different magnitudes of SST warming in the western equatorial Pacific, is associated with differing levels of negative cloud–radiation feedback over the central equatorial Pacific through a dynamic air–sea coupled process involving both the Bjerknes feedback and the wind–evaporation–SST feedback. Considering in isolation the robust common model bias of a weak negative cloud–radiation feedback over the central equatorial Pacific, the projected SST warming in the western equatorial Pacific is likely to be smaller than the multi-model ensemble mean, thereby presenting a more weakeened zonal SST gradient than expected, implying the potential for more severe climate extremes under global warming.

The occurrence of blocking weather patterns over Europe is analysed in a large ensemble of simulations of a climate model with perturbed physical parameters. The experiments were performed with HadGEM3‐GC3 for the UK Climate Change Projections, and comprise a set of 15 coupled simulations supported by a larger suite of 505 atmosphere‐only simulations. Despite the systematic perturbation of 47 different physical constants in the atmosphere‐only experiments, only three were found to have any impact on European blocking frequencies. These reveal the sensitivity of European blocking to orographic drag in winter and to convective entrainment in summer. However, these sensitivities cannot be traced through to the coupled simulations, due to the smaller and more realistic range of perturbations used and likely also to coupled dynamical effects. Overall, we find that although physical sensitivity to the parameterisations exists, adjustment of the parameters is no replacement for further structural improvement in the representation of these processes in the model.

The Diurnal Land–Atmosphere Coupling Experiment (DICE) aims to explore the complex interactions between the land surface and atmospheric boundary layer, which are generally not well understood and difficult to isolate in models. The project involves over 10 different models, combining expertise from both land‐surface and atmospheric boundary‐layer modelling groups. A simple three‐stage methodology is designed to assess land–atmosphere feedbacks. Stage 1: the individual components are assessed in isolation, driven and evaluated against observational data; stage 2: the impact of coupling is investigated; stage 3: the sensitivity of the stand‐alone models to variations in driving data is explored. For this initial study, a 3‐day clear‐sky period in the mid‐west United States over, an assumed simple, predominantly grass surface was simulated using data from the CASES‐99 field campaign. Key conclusions from the study include: (1) the memory of vegetation state within land‐surface models needs attention; (2) the height of atmospheric forcing for land‐surface models is important, particularly for the nocturnal boundary layer, and this has implications for both observations and vertical resolution for atmospheric models; (3) land–atmosphere feedbacks reduce errors in simulated surface fluxes at the expense of the accuracy of the variables that the models are designed to simulate (e.g., temperature, humidity, and wind speed); (4) problems remain in representing the stable boundary layer in atmospheric models; (5) the mixing of temperature and humidity within the boundary layer may need to be represented separately; (6) differences in daytime profiles of heat, moisture, and momentum between models are mainly due to the way the models erode the inversion at the top of the boundary layer, rather than differences in the surface fluxes. Resultant variations in modelled boundary‐layer heights have a substantial impact on relative humidity and could partially explain variations in coupling strength between models in the Global Land–Atmosphere Coupling Experiment.

Tropical land generally experiences the hottest period (spring) in a year just before the onset of wet season. Previous studies suggested that in a warming climate, the wet season would come later, but its origin is debated and its impact on temperature remains unknown. Here, we find that the warming of hot season would be amplified under global warming, and refer to it as “hot-season-gets-hotter” phenomenon. The amplified hot season warming is closely tied to the amplified warming of hot temperature percentiles. The hot-season-gets-hotter phenomenon is mainly due to the rainfall delay and most evident in the Amazon, where spring is warming by almost 1 K more than the annual mean and the 99th percentile temperatures are warming ~30% more than the mean by the end of 21st century in a high emission scenario. Comparing experiments with and without land-atmosphere coupling, it is further found that the rainfall delay is initially driven by the enhanced effective atmospheric heat capacity and then substantially amplified by positive soil moisture-atmosphere feedback. In the satellite period, observations consistently show that the hot-season-gets-hotter phenomenon has already emerged along with the rainfall delay in the Amazon. Intensified hot and dry spring climate can enhance risks of drought, heatwaves and wildfires, threatening the Amazon forest and habitats in the tropics.

Explosive volcanic eruptions can produce large masses of tephra that are transported over long distances, with potential impacts on the ground and to aircraft. Volcanic Ash Advisory Centers (VAACs) provide advice to aviation following an eruption, using atmospheric dispersion models, initialized with eruption source parameters (ESPs) and driven by forecast meteorological data. In this paper, we develop a framework for producing probabilistic forecasts incorporating uncertainty in these inputs. Meteorological uncertainty is typically provided as an ensemble of numerical weather prediction (NWP) data and ESPs include eruption plume height and mass eruption rate (MER); these are linked by atmospheric processes, and their relationship can be modeled by Bayesian regression to quantify their uncertainties. These uncertainties can be propagated through the model to compute probabilistic quantities of ash concentration. The linearity of the advection‐diffusion‐sedimentation (ADS) equation solved by dispersion models allows us to run a single simulation for each of the NWP ensemble members, and then rescale the results to any combination of MER (or height) and emission profile. This gives a computational speed‐up compared to conventional approaches of computing every combination of ESPs and NWP. We demonstrate our method in the operational setting of the London VAAC, using the UK Met Office's NAME dispersion model, although it can be applied to any eruption scenario using any dispersion model that solves the ADS equation. The major source of uncertainty for this case study arises from the MER (due to limited variation in NWP data), although both sources of uncertainty are significant.

Ambient air pollution remains a global challenge, with adverse impacts on health and the environment. Addressing air pollution requires reliable data on pollutant concentrations, which form the foundation for interventions aimed at improving air quality. However, in many regions, including the United Kingdom, air pollution monitoring networks are characterized by spatial sparsity, heterogeneous placement, and frequent temporal data gaps, often due to issues such as power outages. We introduce a scalable data-driven supervised machine learning model framework designed to address temporal and spatial data gaps by filling missing measurements within the United Kingdom. The machine learning framework used is LightGBM, a gradient boosting algorithm based on decision trees, for efficient and scalable modeling. This approach provides a comprehensive dataset for England throughout 2018 at a 1 km ² hourly resolution. Leveraging machine learning techniques and real-world data from the sparsely distributed monitoring stations, we generate 355,827 synthetic monitoring stations across the study area. Validation was conducted to assess the model’s performance in forecasting, estimating missing locations, and capturing peak concentrations. The resulting dataset is of particular interest to a diverse range of stakeholders engaged in downstream assessments supported by outdoor air pollution concentration data for nitrogen dioxide (NO 2 ), Ozone (O 3 ), particulate matter with a diameter of 10 μm or less (PM 10 ), particulate matter with a diameter of 2.5 μm or less PM 2.5 , and sulphur dioxide (SO 2 ), at a higher resolution than was previously possible.

Sea ice thickness (SIT) estimates derived from CryoSat‐2 radar freeboard measurements are assimilated into the Met Office's Forecast Ocean Assimilation Model. We test the sensitivity of winter simulations to the snow depth, radar freeboard product and assumed radar penetration through the snowpack in the freeboard‐to‐thickness conversion. We find that modifying the snow depth has the biggest impact on the modeled SIT, changing it by up to 0.88 m (48%), compared to 0.65 m (33%) when modifying the assumed radar penetration through the snowpack and 0.55 m (30%) when modifying the freeboard product. We find a doubling in the thermodynamic volume change over the winter season when assimilating SIT data, with the largest changes seen in the congelation ice growth. Next, we determine that the method used to calculate the observation uncertainties of the assimilated data products can change the mean daily model SIT by up to 36%. Compared to measurements collected at upward‐looking sonar moorings and during the Operation IceBridge campaign, we find an improvement in the SIT simulations' variability representation when assuming partial radar penetration through the snowpack and when improving the method used to calculate the CryoSat‐2 observation uncertainties. This paper highlights a concern for future SIT data assimilation and forecasting, with the chosen parameterization of the freeboard‐to‐thickness conversion having a substantial impact on model results.

A non-autonomous system can undergo a rapid change of state in response to a small or slow change in forcing, due to the presence of nonlinear processes that give rise to critical transitions or tipping points. Such transitions are thought possible in various subsystems (tipping elements) of the Earth’s climate system. The Atlantic Meridional Overturning Circulation (AMOC) is considered a particular tipping element where models of varying complexity have shown the potential for bi-stability and tipping. We consider both transient and stochastic forcing of a simple but data-adapted model of the AMOC. We propose and test a geometric early warning signal to predict whether tipping will occur for large transient forcing, based on the dynamics near an edge state. For stochastic forcing, we quantify mean times between noise-induced tipping in the presence of stochastic forcing using an Ordered Line Integral Method of Cameron (2017) to estimate the quasipotential. We calculate minimum action paths between stable states for various scenarios. Finally, we discuss the problem of finding early warnings in the presence of both transient and stochastic forcing.

In numerical weather prediction (NWP), a large number of observations are used to create initial conditions for weather forecasting through a process known as data assimilation. An assessment of the value of these observations for NWP can guide us in the design of future observation networks, help us to identify problems with the assimilation system, and allow us to assess changes to the assimilation system. However, assessment can be challenging in convection‐permitting NWP. This is because verification of convection‐permitting forecasts is not easy, the forecast model is strongly nonlinear, a limited‐area model is used, and the observations used often contain complex error statistics and are often associated with nonlinear observation operators. We compare methods that can be used to assess the value of observations in convection‐permitting NWP and discuss operational considerations when using these methods. We focus on their applicability to ensemble forecasting systems, as these systems are becoming increasingly dominant for convection‐permitting NWP. We also identify several future research directions, which include comparing results from different methods, comparing forecast validation using analyses versus using observations, applying flow‐dependent covariance localization, investigating the effect of ensemble size on the assessment, and generating and validating the nature run in observing‐system simulation experiments.